Line array laser push-broom system failure detection and bidirectional switching time gating method

Abstract

This invention discloses a failure detection and bidirectional switching time-gated method for a linear laser push-broom system. By designing a multi-feature evidence body under range-gated detection mode and introducing D-S evidence theory to achieve multi-dimensional feature evidence fusion, it effectively solves the problem of low accuracy in traditional single-parameter threshold failure detection. A closed-loop switching mechanism is designed to revert to a wide window under range-gated narrow-window failure conditions. Based on the system state judgment result and the prior value of the target distance, a wide-window restart and parameter reset are triggered after narrow-window failure, forming a bidirectional switching closed-loop control of wide-window target search – narrow-window precise detection – failure reversion. This invention offers high detection accuracy, strong anti-interference capability, and adapts to the real-time requirements of multi-channel parallel processing in linear arrays. It effectively avoids the impact of blind elements, crosstalk, and other failures on imaging quality, significantly improving the operational reliability of the linear laser push-broom system.

Description

Technical Field

[0001] This invention belongs to the field of laser detection and photoelectric imaging, specifically relating to a failure detection and bidirectional switching time gating method for a linear laser pushbroom system. Background Technology

[0002] Linear laser pushbroom systems are widely used in lidar, topographic mapping, and guided detection due to their advantages such as multi-channel parallel detection and high-resolution imaging. Their core detection unit is a linear APD array, which achieves high-precision ranging over a wide range through a range gating mode of wide-window search and narrow-window precision measurement. However, the range gating interval is usually narrow, and traditional fixed-step scanning is prone to target information loss due to mismatch between the narrow window and the target position, leading to detection failure and reducing the system's target detection capability.

[0003] Currently, scholars have conducted extensive research to address the problem of range-gated detection failure and have proposed many methods for adjusting and controlling the range-gated window. Sun Junquan et al. proposed a multi-pulse time-series coded gated range-gated imaging method based on morphological operations. By introducing morphological opening operations, the system's dependence on the binarization threshold during decoding can be effectively reduced, thereby improving target integrity (Sun Junquan, Yin Yongkai, Wei Yifan, et al. Gated range-gated three-dimensional imaging based on morphological operations [J]. Optoelectronic Technology Application, 2025, 40(04): 1-7.). Chen Songmao et al. proposed a target information extraction and three-dimensional reconstruction method under wide target detection range conditions. Through a range-gated algorithm based on the target information distribution range, high-precision three-dimensional reconstruction is achieved, which has stronger noise suppression capabilities than pure hardware-based denoising methods (Chen Songmao, Su Xiuqin, Hao Wei, et al. Adaptive gated noise suppression and three-dimensional reconstruction algorithm based on photon counting lidar [J]. Acta Physica Sinica, 2022, 71(10): 222-233.).

[0004] Existing range gating control methods mostly rely on threshold conditions with a single parameter to trigger the switching of detection modes. These methods are ineffective in addressing issues such as target loss and detection failure due to range gating window malfunctions caused by target distance offset or amplifier saturation. Therefore, this invention proposes a failure detection and bidirectional switching time-gating method for a linear laser pushbroom system to solve these problems. Summary of the Invention

[0005] This invention proposes a failure detection and bidirectional switching time gating method for a linear laser push-broom system, which effectively solves the problem of functional failure in the distance-gated detection mode of the linear laser push-broom system and improves the detection robustness in failure scenarios.

[0006] The technical solution for achieving the present invention is as follows: a time-gated method for failure detection and bidirectional switching of a linear laser push-broom system, comprising the following steps:

[0007] Step 1: Construct the linear laser push-broom system and the multi-dimensional failure detection and control module respectively:

[0008] The linear laser push-broom system includes a laser drive control module, a linear laser emitting module, an emitting optics module, a receiving optics module, a linear APD receiving module, a front-end readout module, and a range gating module. The laser drive control module, the linear laser emitting module, and the emitting optics module are connected in sequence. The receiving optics module, the linear APD receiving module, and the front-end readout module are connected in sequence. The range gating module is connected to the linear APD receiving module.

[0009] The multidimensional failure discrimination and control module includes a main control unit, a failure discrimination module, a failure compensation module, and a data processing and status monitoring unit. The failure discrimination module is connected to the main control unit, the failure compensation module, the distance gating module, and the data processing and status monitoring unit. The data processing and status monitoring unit is connected to the front-end readout module. The main control unit is connected to the linear APD receiving module, the failure compensation module, and the laser drive control module.

[0010] Proceed to step 2.

[0011] Step 2: During the system initialization phase, the main control unit sends a set of standard test pulses and records the baseline voltage of each channel. Proceed to step 3.

[0012] Step 3: The main control unit determines the detection range based on the preset parameters. , set the first Initial gating window parameters for each channel Proceed to step 4.

[0013] Step 4: During the detection process of the linear laser push-broom system, the data processing and status monitoring unit monitors the echo signal of each channel of the linear laser push-broom system in real time and extracts 6 failure characteristic parameters, namely signal-to-noise ratio, effective count rate, false alarm rate, baseline drift, signal amplitude linearity deviation, and blocking time, and then proceeds to step 5.

[0014] Step 5: In the failure determination module, a multi-dimensional feature fusion failure determination algorithm based on DS evidence theory is designed. Six failure feature parameters are used as evidence to perform multi-dimensional failure determination on each channel of the linear laser push-broom system, output the system status determination result, monitor the detection status of each channel in real time, and then proceed to step 6.

[0015] Step 6: Based on the system status determination results, design a discrete event-driven logic control system based on status feedback to realize the full-link closed-loop control of the linear laser push-broom system for wide-window search, narrow-window precision measurement, failure backoff, and re-searching for the target. When the system status determination result is normal, maintain the system detection state; when the system status determination result is failure, if the system is in wide-window detection state, maintain the detection mode unchanged, adjust the detection window width, and re-acquire the target; if it is in narrow-window detection mode, trigger the switching from narrow-window to wide-window detection mode, return to wide-window detection mode, reset the initial parameters of the wide window, and compensate the blind element channel of the linear laser push-broom system through the failure compensation module.

[0016] Compared with the prior art, the significant advantages of this invention are:

[0017] (1) This invention designs a multi-feature evidence body under the distance gating detection mode and introduces DS evidence theory to complete the fusion of multi-dimensional feature evidence, which effectively improves the system failure discrimination accuracy and anti-interference ability.

[0018] (2) Based on the system state determination results of DS evidence fusion, the present invention designs a closed-loop control logic of wide window search, narrow window precision measurement, failure backoff, and re-search for the target. When the narrow window detection fails, the system triggers the narrow window to return to the wide window and resets the gating parameters to achieve continuous tracking and detection of the target.

[0019] (3) In view of the characteristics of multi-channel parallel detection in linear laser push-broom system, the present invention realizes independent failure judgment and mode control for each channel. Combined with channel interpolation compensation logic, it avoids interference of failed channels on the overall imaging and improves the imaging quality of the system. Attached Figure Description

[0020] Figure 1 This is the overall flowchart of the present invention.

[0021] Figure 2 The present invention relates to a linear laser push-broom system.

[0022] Figure 3 This is a diagram of the overall architecture of the present invention.

[0023] Figure 4 This is the DS multi-dimensional feature fusion process of the present invention.

[0024] Figure 5 This is the evidence interval after feature fusion in this invention. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0026] Furthermore, the technical solutions of the various embodiments of the present invention can be combined with each other, but only if they are feasible to those skilled in the art. If the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such combination of technical solutions does not exist and is not within the scope of protection claimed by the present invention.

[0027] The following section will further introduce the specific implementation method, as well as the technical difficulties and inventive points of this invention, using this design example as an example.

[0028] Combination Figure 1 The present invention discloses a method for failure detection and bidirectional switching time gating of a linear laser push-broom system, the implementation steps of which are as follows:

[0029] Step 1: Construct the linear laser push-broom system and the multi-dimensional failure detection and control module respectively:

[0030] The linear array laser pushbroom system includes a laser drive control module, a linear array laser emitting module, an emitting optics module, a receiving optics module, a linear array APD receiving module, a front-end readout module, and a range gating module. The laser drive control module, the linear array laser emitting module, and the emitting optics module are connected in sequence. The receiving optics module, the linear array APD receiving module, and the front-end readout module are connected in sequence. The range gating module is connected to the linear array APD receiving module. The laser drive control module, the range gating module, and the linear array APD receiving module are directly controlled by a multi-dimensional failure discrimination control module. The front-end readout module and the range gating module respectively feed back real-time laser detection data and range gating status of the linear array laser pushbroom system to the multi-dimensional failure discrimination control module.

[0031] The multi-dimensional failure discrimination and control module includes a main control unit, a failure discrimination module, a failure compensation module, and a data processing and status monitoring unit. The failure discrimination module is connected to the main control unit, the failure compensation module, the range gating module, and the data processing and status monitoring unit. The data processing and status monitoring unit is connected to the front-end readout module. The main control unit is connected to the linear array APD receiving module, the failure compensation module, and the laser drive control module. The failure discrimination module and the data processing and status monitoring unit receive data output from the range gating module and the front-end readout module, respectively. The main control unit and the failure compensation module are responsible for controlling and compensating the linear array laser push-broom system.

[0032] like Figure 2 After the linear array laser push-broom system starts working, the main control unit sends pulse drive signals to the laser drive control module according to the movement speed of the carrier and the preset frame rate. After receiving the laser drive signal, each laser emitting unit in the linear array laser emitting module generates several laser beams in sequence. The emitted laser beams are collimated and focused by the emitting optical module and then emitted towards the target, forming a laser line array with a certain field of view in the target area.

[0033] like Figure 3 The receiving optical module filters out stray light from the echo signal and inputs the valid echo signal into the linear APD receiving module. The main control unit sequentially enables each receiving unit in the linear APD receiving module via an enable signal. The distance gating control module controls each receiving unit to receive the echo signal within a preset gate window and converts the received optical signal into a current signal. The front-end readout module converts the current signal into a voltage signal. The data processing and status detection unit calculates the time interval between the laser drive signal and the echo signal and obtains laser detection data for several channels based on the pulse time-of-flight method. This data includes six failure characteristic parameters: laser echo signal, laser ranging result, signal-to-noise ratio, effective count rate, false alarm rate, signal amplitude linearity deviation, baseline drift, and blocking time. The system detection status is then determined and controlled using this data.

[0034] Step 2: During the system initialization phase, the main control unit sends a set of standard test pulses and records the baseline voltage of each channel. This provides reference data for the calculation of the subsequent six failure characteristic parameters.

[0035] Step 3: When the system is working, the main control unit will determine the detection range based on the preset parameters. , set the first Initial gating window parameters for each channel :

[0036] (1),

[0037] in, Represents the speed of light. Indicates the first The preset minimum distance for each channel. Indicates the first The maximum distance preset for each channel This indicates the door protection time, providing a certain safety margin for the door-controlled window.

[0038] Once the system obtains stable laser detection data, the ranging result of the previous frame in each channel is used as the prediction center. Constructing a narrow window in the distance domain Narrow window for distance domain Mapping to the time domain yields a narrow time window. The main control unit sends a range gating control signal to the range gating module based on the narrow window data of each channel, converging the effective detection window of the linear APD receiving module to a narrow window near the predicted detection result. To reduce the probability of false triggering of the linear laser push-broom system due to environmental noise, a narrow window in the range domain is used. and time domain narrow window The calculation formula is:

[0039] (2),

[0040] (3),

[0041] in, This indicates the upper bound of the prediction error.

[0042] Step 4: During the detection process of the linear laser push-broom system, the data processing and status monitoring unit monitors the echo signal of each channel of the linear laser push-broom system in real time and extracts six failure characteristic parameters: signal-to-noise ratio, effective count rate, false alarm rate, baseline drift, signal amplitude linearity deviation, and blocking time.

[0043] System signal-to-noise ratio The ability to characterize the real echo signal from noise directly reflects the range detection accuracy. The main control unit calculates the noise mean by extracting the echo signals from the noise region and the signal region respectively based on the gating control signal. noise energy and signal energy And calculate the system signal-to-noise ratio. The noise energy, signal energy, and system signal-to-noise ratio are calculated as follows:

[0044] (4),

[0045] (5),

[0046] in These represent the laser echo signals in the noise region and the signal region, respectively.

[0047] Peak detection is performed using the front-end readout module, and the number of valid echoes at different gating distances per unit time is recorded as the system's effective count rate. The number of spikes exceeding the threshold in the noise zone per unit time is recorded as the false alarm rate. Baseline drift This represents the DC output level of the APD in linear operating mode without light, directly reflecting the APD's dark-state stability. Excessive drift can cause signal threshold determination to fail. The calculation method is as follows:

[0048] (6),

[0049] This represents the baseline voltage under the current condition. This represents the baseline voltage calibrated during system initialization. Signal amplitude linearity deviation. The deviation between the actual output amplitude and the ideal output amplitude of the APD is reflected and can be calculated by the difference between the current channel's output amplitude and the average amplitude of neighboring channels. Blocking time. This indicates the time it takes for the front-end readout module to recover from saturation to the normal linear operating range, obtained through the data processing and status monitoring unit. In a linear laser push-broom system, to avoid ranging data corruption caused by channel crosstalk, the blocking time of each channel needs to be strictly matched with the timing cycle of channel switching.

[0050] Step 5: To address the issues of fixed linear weights and weak anti-interference capability in traditional weighted summation fusion algorithms, a multi-dimensional feature fusion failure determination algorithm based on DS evidence theory is designed in the failure determination module. This algorithm uses the system's six failure feature parameters as evidence to perform multi-dimensional failure determination on each channel of the linear laser push-broom system, monitoring the detection status of each channel in real time to ensure the quality of the output point cloud. This method has a relatively low computational load and can adapt to the lightweight and real-time requirements of linear laser push-broom systems. The specific steps are as follows:

[0051] Based on the system failure classification requirements, a system identification framework is defined:

[0052] (7),

[0053] This indicates the detection status of each channel in the linear laser push-broom system, including normal. With failure Two types.

[0054] Define the power set:

[0055] (8),

[0056] Representation of recognition framework The set of all subsets.

[0057] because Includes normal With failure Since there are two states, the allocation cannot be determined. False alarm rate, signal amplitude linearity deviation, blocking time, and baseline drift are defined as positive parameters; the larger the parameter values, the greater the risk of system failure. Signal-to-noise ratio and effective count rate are defined as negative parameters; the smaller the parameter values, the greater the risk of system failure. The basic probability allocation function for the six evidence bodies in the system is initialized based on the characteristics of these different parameters. ,function The following conditions must be met:

[0058] (9),

[0059] In this context, any event in the power set... .

[0060] For each failure characteristic parameter, its normal operating range and failure critical range are determined through offline calibration experiments and historical operating data analysis. Based on these ranges, membership functions for each failure characteristic parameter with respect to system normal operation and failure are constructed. The boundary threshold parameters of the membership functions are then embedded into the main control unit as initialization parameters for the basic probability allocation function. .

[0061] During the operation of the online laser push-broom system, six failure characteristic parameters are collected in real time for each detection channel: signal-to-noise ratio, effective count rate, false alarm rate, signal amplitude linearity deviation, baseline drift, and blocking time. After normalization, these parameters are used as a set of evidence as shown in formula (10) and input into the failure discrimination module.

[0062] (10)

[0063] The basic probability assignment value of each failure characteristic parameter for the normal and failure states of the detection channel is calculated based on the pre-calibrated membership function. The calculation method is as follows:

[0064] (11),

[0065] in , Threshold conditions representing system failure and normal states. , This represents the basic probability assignment of the current failure characteristic parameter value to the system state. It represents the probability that the system state is uncertain. Represents the set of evidence The evidence in the middle, .

[0066] like Figure 4When multiple parameters provide different support for the system state, the Dempster combination rule is used to calculate the comprehensive basic probability allocation of each parameter for the system state, as expressed in the following formula:

[0067] (12)

[0068] This represents the conflict coefficient between failure characteristic parameters. A larger value indicates a more severe contradiction between different failure characteristic parameters. This represents the normalization factor, used to balance the conflicts between failure characteristic parameters. , Represents the power set Any event in, This indicates that the first piece of evidence relates to the event. The basic probability distribution that holds true. This indicates that the second piece of evidence relates to the event. The basic probability distribution of the two evidence bodies is calculated using formula (12) for all intersections. The comprehensive basic probability allocation of a subset of the system is obtained by using six failure characteristic parameters as evidence bodies and sequentially fusing the comprehensive basic probability allocations of the six evidence bodies for the three detection states of the system. , and This serves as the fundamental basis for determining the system's detection status.

[0069] To further reduce the risk of system misjudgment, a trust function Bel and a likelihood function Pl are defined for system failure, and an evidence interval for the system failure proposition is established as follows: Figure 5 As shown, The confidence interval for the system's judgment result is represented by the following formula: The confidence function Bel and the likelihood function Pl are calculated for the system failure and normal states, respectively.

[0070] (13)

[0071] Since the two states of the linear laser push-broom system are independent of each other, the trust function Bel for the failure and normal states is equal to the comprehensive basic probability allocation, as shown in formula (14):

[0072] (14)

[0073] The confidence interval of the system's discrimination result is The smaller the confidence interval, the lower the uncertainty of the judgment result, and the more reliable the judgment result. The candidate judgment result with the larger confidence function value between the two states is selected. If the corresponding uncertainty is less than the threshold, the system state judgment result is output; otherwise, the system state is re-judged.

[0074] Step 6: Based on the system state determination results, design a discrete event-driven logic control system based on state feedback to realize the full-link closed-loop control of the linear laser push-broom system for wide-window search, narrow-window precision measurement, failure backoff, and re-search for the target.

[0075] Define the system probe state transition function as follows:

[0076] (15)

[0077] (16)

[0078] in, This indicates the system's current detection mode. , These represent the system's wide-window capture state and narrow-window precise detection state, respectively. This indicates the system status determination result output by the failure detection module, where 0 represents a normal state and 1 represents a detection failure. It also represents the width of the detection window for the next system time step. Determined by the current detection mode and detection status, , These represent the detection window widths for wide-window and narrow-window detection modes, respectively.

[0079] When the system status is determined to be normal, the system detection mode is maintained. When the system status is determined to be faulty, if the system is in wide-window detection mode, the detection mode remains unchanged, the detection window width is adjusted, and the target is reacquired; if it is in narrow-window detection mode, a switch from narrow-window to wide-window detection mode is triggered. After returning to wide-window detection mode, the initial parameters of the wide window are reset according to the target distance prediction range to shorten the time for reacquiring the target, and the faulty channel is marked to avoid channel crosstalk causing ranging errors. The failure compensation module compensates for the linear laser pushbroom system to reduce the impact of the faulty channel on the overall imaging quality.

[0080] Channels that are deemed invalid for 5 consecutive frames are identified as blind channels. , Channel number, The sequence number is the pushbroom frame number. Valid data within the spatiotemporal neighborhood of the blind pixel channel is filtered, where the temporal neighborhood consists of adjacent frames above and below each other in the same channel, and the spatial neighborhood consists of adjacent channels to the left and right within the same frame. If valid laser detection data exists in both the spatiotemporal neighborhoods of the blind pixel channel, bilinear interpolation is used to calculate the compensation value. The compensation value replaces the laser detection data in the blind pixel channel and participates in subsequent 2D point cloud image stitching. Interpolation compensation is performed in real time after each frame pushbroom to ensure the real-time performance of the system. The compensation result is dynamically updated according to the blind pixel channel status. If the channel returns to normal, compensation stops and the original laser detection data is restored.

[0081] The above bilinear interpolation method is used to calculate the compensation for laser detection data in the blind cell channel as shown in equation (17):

[0082] (17)

[0083] in, Indicates the first The first channel, the... Frame of laser detection data.

[0084] To avoid frequent switching of detection modes, an anti-shake strategy is designed. After a mode switch is completed, a timer is started, and the system does not respond to another switch request within 5 frames.

[0085] The innovation of this invention lies in proposing a failure detection and bidirectional switching time gating method for a linear laser pushbroom system. Based on the DS evidence theory, a wide-window closed-loop control system for failure detection of distance gating in the linear laser pushbroom system is constructed. This solves the problems of missed detection and misjudgment caused by target distance changes, gating window mismatch, and environmental noise coupling in traditional single-parameter threshold detection, effectively improving the accuracy of system failure detection. Furthermore, a blind element channel compensation method using bilinear interpolation is designed to ensure the point cloud quality output by the linear laser pushbroom system.

Claims

1. A method for failure detection and bidirectional switching time gating of a linear laser push-broom system, characterized in that, The steps are as follows: Step 1: Construct the linear laser push-broom system and the multi-dimensional failure detection and control module respectively: The linear array laser push-broom system includes a laser drive control module, a linear array laser emitting module, an emitting optical module, a receiving optical module, a linear array APD receiving module, a front-end readout module, and a range gating module. The laser drive control module, the linear array laser emitting module, and the emitting optical module are connected in sequence. The receiving optical module, the linear array APD receiving module, and the front-end readout module are connected in sequence. The range gating module is connected to the linear array APD receiving module. The multi-dimensional failure discrimination and control module includes a main control unit, a failure discrimination module, a failure compensation module, and a data processing and status monitoring unit. The failure discrimination module is connected to the main control unit, the failure compensation module, the distance gating module, and the data processing and status monitoring unit. The data processing and status monitoring unit is connected to the front-end readout module. The main control unit is connected to the linear APD receiving module, the failure compensation module, and the laser drive control module. Proceed to step 2; Step 2: During the system initialization phase, the main control unit sends a set of standard test pulses and records the baseline voltage of each channel. Proceed to step 3; Step 3: The main control unit determines the detection range based on the preset parameters. , set the first Initial gating window parameters for each channel Proceed to step 4; Step 4: During the detection process of the linear laser push-broom system, the data processing and status monitoring unit monitors the echo signal of each channel of the linear laser push-broom system in real time and extracts 6 failure characteristic parameters, namely signal-to-noise ratio, effective count rate, false alarm rate, baseline drift, signal amplitude linearity deviation, and blocking time, and then proceeds to step 5. Step 5: In the failure determination module, a multi-dimensional feature fusion failure determination algorithm based on DS evidence theory is designed. Six failure feature parameters are used as evidence to perform multi-dimensional failure determination on each channel of the linear laser push-broom system, output the system status determination result, monitor the detection status of each channel in real time, and proceed to step 6. Step 6: Based on the system state determination results, design a discrete event-driven logic control system based on state feedback to realize the full-link closed-loop control of the linear laser push-broom system for wide-window search, narrow-window precision measurement, failure backoff, and re-search for the target. If the system status determination result is normal, maintain the system detection state; When the system status is determined to be faulty, if the system is in wide-window detection mode, the detection mode remains unchanged, the detection window width is adjusted, and the target is re-acquired; if it is in narrow-window detection mode, the detection mode switching from narrow window to wide window is triggered, the system returns to wide-window detection mode, the initial parameters of the wide window are reset, and the blind element channel of the linear laser push-broom system is compensated through the failure compensation module.

2. The failure detection and bidirectional switching time-gated method for a linear laser push-broom system according to claim 1, characterized in that, In step 1, after the linear array laser push-broom system starts working, the main control unit sends a pulse drive signal to the laser drive control module according to the movement speed of the carrier and the preset frame rate. After receiving the laser drive signal, each laser emitting unit in the linear array laser emitting module generates several laser beams in sequence. The emitted laser beams are collimated and focused by the emitting optical module and then emitted towards the target, forming a laser line array with a certain field of view in the target area. The receiving optical module filters out stray light from the echo signal and inputs the valid echo signal into the linear APD receiving module. The main control unit enables each receiving unit in the linear APD receiving module sequentially through the enable signal. The distance gating control module controls each receiving unit to receive the echo signal within the preset gate window and converts the received optical signal into a current signal. The front-end readout module converts the current signal into a voltage signal. The data processing and status detection unit calculates the time interval between the laser drive signal and the echo signal and obtains laser detection data for several channels according to the pulse time-of-flight method, including six failure characteristic parameters: laser echo signal, laser ranging result, signal-to-noise ratio, effective count rate, false alarm rate, signal amplitude linearity deviation, baseline drift, and blocking time. The above data is used to determine and control the system detection status.

3. The failure detection and bidirectional switching time-gating method for a linear laser push-broom system according to claim 2, characterized in that, In step 3, when the system is working, the main control unit operates according to the preset detection range. , set the first Initial gating window parameters for each channel The details are as follows: （1）， in, Represents the speed of light. Indicates the first The preset minimum distance for each channel. Indicates the first The maximum distance preset for each channel This indicates the door protection time, providing a certain safety margin for door-controlled windows; Once the system obtains stable laser detection data, the ranging result of the previous frame in each channel is used as the prediction center. Constructing a narrow window in the distance domain Narrow window for distance domain Mapping to the time domain yields a narrow time window. The main control unit sends a range gating control signal to the range gating module based on the narrow window data of each channel, converging the effective detection window of the linear APD receiving module to a narrow window near the predicted detection result. To reduce the probability of false triggering of the linear laser push-broom system due to environmental noise, a narrow window in the range domain is used. and time domain narrow window The calculation formula is: （2）， （3）， in, This indicates the upper bound of the prediction error.

4. The failure detection and bidirectional switching time-gating method for a linear laser push-broom system according to claim 3, characterized in that, Step 4 is detailed below: The main control unit extracts the echo signals from the noise region and the signal region respectively based on the gating control signal and calculates the noise mean. noise energy and signal energy And calculate the system signal-to-noise ratio. Noise energy, signal energy, and system signal-to-noise ratio; Peak detection is performed using the front-end readout module, and the number of valid echoes at different gating distances per unit time is recorded as the system's effective count rate. The number of spikes exceeding the threshold in the noise zone per unit time is recorded as the false alarm rate. ; Baseline drift This represents the DC output level of the APD in linear operating mode without light, directly reflecting the APD's dark-state stability. Excessive drift can cause signal threshold determination to fail. The calculation method is as follows: （6）， in, This represents the baseline voltage under the current condition. This indicates the baseline voltage calibrated during system initialization; Signal amplitude linearity deviation Reflecting the deviation between the actual output amplitude and the ideal output amplitude of the APD, it is calculated by the difference between the current channel output amplitude and the average amplitude of the neighboring channels; blocking time. This indicates the time it takes for the front-end readout module to recover from saturation to the normal linear operating range, which is obtained through the data processing and status monitoring unit. In the linear laser push-broom system, in order to avoid the ranging data disorder caused by channel crosstalk, the blocking time of each channel needs to be strictly matched with the timing cycle of channel switching.

5. The failure detection and bidirectional switching time-gated method for a linear laser push-broom system according to claim 4, characterized in that, Step 5 is detailed below: Based on the system failure classification requirements, a system identification framework is defined: （7）， in, This indicates the detection status of each channel in the linear laser push-broom system, including normal. With failure Two kinds; Define the power set: （8）， Representation of recognition framework The set of all subsets; False alarm rate, signal amplitude linearity deviation, blocking time, and baseline drift are defined as positive parameters. The larger the parameter value, the greater the risk of system failure. Signal-to-noise ratio and effective count rate are defined as negative parameters. The smaller the parameter value, the greater the risk of system failure. Initialize the basic probability allocation function of the six pieces of evidence in the system based on different parameter characteristics. ,function The following conditions must be met: （9）， In this context, any event in the power set... ; For each failure characteristic parameter, its normal operating range and failure critical range are determined through offline calibration experiments and historical operating data analysis. Based on these ranges, membership functions for each failure characteristic parameter with respect to system normal operation and failure are constructed. The boundary threshold parameters of the membership functions are then embedded into the main control unit as initialization parameters for the basic probability allocation function. ; During the operation of the online laser push-broom system, six failure characteristic parameters are collected in real time for each detection channel: signal-to-noise ratio, effective count rate, false alarm rate, signal amplitude linearity deviation, baseline drift, and blocking time. After normalization, these parameters are used as a set of evidence as shown in formula (10) and input into the failure discrimination module. （10）， The basic probability assignment value of each failure characteristic parameter for the normal and failure states of the detection channel is calculated based on the pre-calibrated membership function. The calculation method is as follows: （11）， in, , Threshold conditions representing system failure and normal states. , This represents the basic probability assignment of the current failure characteristic parameter value to the system state. It represents the probability that the system state is uncertain; Represents the set of evidence The evidence in the middle, ; When multiple parameters provide different support for the system state, the Dempster combination rule is used to calculate the comprehensive basic probability allocation of each parameter for the system state, as expressed in the following formula: （12）， This represents the conflict coefficient between failure characteristic parameters. A larger value indicates a more severe contradiction between different failure characteristic parameters. This represents the normalization factor, used to balance conflicts between failure characteristic parameters; , Represents the power set Any event in, This indicates that the first piece of evidence relates to the event. The basic probability distribution that holds true. This indicates that the second piece of evidence relates to the event. The basic probability distribution of the two evidence bodies is calculated using formula (12) for all intersections. The comprehensive basic probability assignment of a subset of ; Using six failure characteristic parameters as evidence, the comprehensive basic probability allocation of the system's three detection states is obtained by sequentially fusing the six evidences. , and This serves as the fundamental basis for determining the system's detection status. To further reduce the risk of system misjudgment, a trust function Bel and a likelihood function Pl are defined for system failure, and an evidence interval for the system failure proposition is established. The confidence interval for the system's judgment result is represented by the following formula: The confidence function Bel and the likelihood function Pl are calculated for the system failure and normal states, respectively. （13）， Since the two states of the linear laser push-broom system are independent of each other, the trust function Bel for the failure and normal states is equal to the comprehensive basic probability allocation, as shown in formula (14): （14）， The confidence interval of the system's discrimination result is The smaller the confidence interval, the smaller the uncertainty of the judgment result, and the more reliable the judgment result. The larger confidence function value of the two states is taken as the candidate judgment result. If the corresponding uncertainty is less than the threshold, the system state judgment result is output; otherwise, the system state is re-judged.

6. The failure detection and bidirectional switching time-gated method for a linear laser push-broom system according to claim 5, characterized in that, In step 6, a discrete event-driven logic control system based on state feedback is designed based on the system state determination results to realize the full-link closed-loop control of the linear laser push-broom system, including wide-window search, narrow-window precision measurement, failure backoff, and re-search for the target. The details are as follows: Define the system probe state transition function as follows: （15）， （16）， in, This indicates the system's current detection mode. , These represent the system's wide-window capture state and narrow-window precise detection state, respectively. This indicates the system status determination result output by the failure detection module, where 0 represents a normal state and 1 represents a detection failure. It also represents the width of the detection window for the next system time step. Determined by the current detection mode and detection status, , These represent the detection window widths for wide-window and narrow-window detection modes, respectively.

7. The failure detection and bidirectional switching time-gated method for a linear laser push-broom system according to claim 6, characterized in that: Channels that are deemed invalid for 5 consecutive frames are identified as blind channels. , Channel number, The pushbroom frame sequence number is used to filter valid data within the spatiotemporal neighborhood of the blind pixel channel. The temporal neighborhood consists of adjacent frames above and below each other in the same channel, and the spatial neighborhood consists of adjacent channels to the left and right within the same frame. If valid laser detection data exists in both the spatiotemporal neighborhoods of the blind pixel channel, bilinear interpolation is used to calculate the compensation value. The compensation value replaces the laser detection data of the blind element channel and participates in the subsequent two-dimensional point cloud image stitching; Interpolation compensation is performed in real time after each frame push-broom is completed to ensure the real-time performance of the system. The compensation result is dynamically updated with the blind cell channel status. If the channel returns to normal, the compensation is stopped and the original laser detection data is restored. The above bilinear interpolation method is used to calculate the compensation for laser detection data in the blind cell channel as shown in equation (17): （17）， in, Indicates the first The first channel, the... Frame of laser detection data; To avoid frequent switching of detection modes, an anti-shake strategy is designed. After a mode switch is completed, a timer is started, and the system does not respond to another switch request within 5 frames.

Images